Weather Characterization

Abstract

The Virginia Tech Transportation Institute (VTTI) successfully developed and deployed a mobile weather characterization system aimed at enhancing transportation safety research at the Virginia Smart Roads facility. This sensor was used to characterize a sampling of the water-based, VTTI simulated weather at the Virginia Smart Roads Facility. A Parsivel2 disdrometer was mounted on a vehicle to measure precipitation particle size distribution and falling velocity. The mobile nature of the system enables efficient data collection along the entire roadway section. Using the sensor, a rain characterization study revealed that the rain produced by the facility showed variability in droplet size distribution, with deviations from natural rain patterns. The limited fall height (10 meters) led to lower terminal velocities than naturally occurring rainfall, which usually fits the Gunn-Kinzer relationship. With respect to the Marshall-Palmer relationship, the VTTI rain represents stratiform rain distribution more than convective rain. Wind was found to have a bigger effect on measurement accuracy due to the sensitivity of the sensor. A snow characterization study revealed challenges in correlating liquid water equivalents measured to actual snow depth due to variability in snow density and particle orientation of the VTTI-produced, water-based snow. The disdrometer software assumes the snow density to calculate the liquid water equivalent. The addition of a heated precipitation gauge could enhance accuracy. Operationally, the study found that calibrating weather towers by pressure, rather than visual estimation, improved the consistency of rainfall production. However, issues such as hose kinks impacted flow rates, indicating areas for infrastructure improvements. Recommendations for future work include enhancements such as wind sensors, articulating mounts, and longer duration testing under various wind conditions are recommended to improve weather characterization fidelity.